TrueML Innovation & Technology Culture

Describe a project you’re especially eager to tackle in the new year.

I’m especially excited to take the Riverty work we completed and evolve it into a fully standardized “European Creditor Ingestion and Reconciliation Framework.” We proved we can handle complex files end to end, and now I want to scale that into something repeatable across multiple portfolios.

The project starts simply: giving teams a consistent, reliable way to upload creditor files, validate them and generate clean placement, adjustment and reconciliation outputs. As the work deepens, I want to make the entire ingestion process more automated and less dependent on manual checks. Technically, this means transforming our current Riverty mapper into a modular pipeline that can parse debt/contract/claims/credit structures, enforce strict validation rules, run waterfall credit allocation, detect duplicates and automatically reconcile new files against historical placements to prevent double-applying credits. Ultimately, I want to build a reusable ingestion template that significantly reduces onboarding time for any European creditor.

What technologies and/or practices is your team leveraging to tackle this project?

We’ll approach the project with a mix of accessible tooling and deeper engineering infrastructure. At a high level, the team will rely on clear documentation, iterative testing with operations and tight collaboration to refine each creditor’s mapping and logic. We’ll continue using sample data to validate assumptions early and avoid rework.

On the technical side, the project will use Retool for the self-serve UI where operations can upload files and review validation results. The mapper and recon logic will be written in JavaScript and Python, leveraging PapaParse, JSZip and Pandas for parsing, transformations and file generation. Snowflake will serve as our system of record for placements, CRED events, historical runs and recon tables. We’ll use Jenkins jobs, or an equivalent continuous integration orchestrator, for scheduling automated ingestion flows. We’ll also incorporate improved logging, error-flagging, duplicate detection and pre-ingestion reconciliation as standard practices across all new creditors.

How does this project tie into larger company goals?

This project supports broader company goals by reducing operational overhead, improving data trust and accelerating new client activations. Starting from a simple goal — make file ingestion smoother — it has grown into something that directly impacts efficiency and client satisfaction.

As the pipeline becomes standardized, onboarding new creditors becomes faster and less risky. Cleaner data leads to more accurate treatment strategies, fewer consumer issues and more predictable recoveries. Automating checks like duplicate invoices, credit-application prevention and placement validation reduces the likelihood of balance disputes and ensures consistent processing across portfolios.

From a technical alignment standpoint, the framework helps scale our ingestion capabilities without proportionally increasing engineering lift. It turns custom creditor logic into a productized, reusable asset that enhances reliability, reduces turnaround time and directly supports revenue growth by enabling us to take on more portfolios with higher confidence.